Goal of this bachelor thesis is to find ways of dynamically detecting player preferences during runtime and to suggest ways of adapting the game according to it.

To do so, it first investigates five different definitions of play and discusses them. It proceeds with a literature search of possible player type taxonomies and discusses them under the aspect of suitability for dynamic detection. The results of this discussion deem Bartle's player types to be the best fit for this task.

The thesis proceeds with outlining a short implementation of a player-type-detection algorithm in a prototype and highlighting possible ways of adapting the game to the players preferences.